Gregory's Server@atomicpoet At least one publication reached out to me to verify the number. Unfortunately I cannot verify the count from the @mastodonusercount bot; fetching total user numbers from all currently existing Mastodon servers using my own algorithm brings me to only 6.5M in comparison.
|
10 comments
 @Gargron @mastodonusercount Thatβs definitely another consideration. For example, Switter is dead and that had over 100K accounts (last I checked). @Gargron @mastodonusercount But that brings your count of 1.3M MAUs into starker focus because if your algorithm is only seeing 6.5M βliveβ accounts, thatβs pretty good retention. @Gargron @mastodonusercount  @atomicpoet @mastodonusercount I likewise suspect that is the reason for the difference, but of course it's hard to confirm.  @Gargron @atomicpoet @mastodonusercount Is it possible that accounts moving from server to server could be counted as duplicates? Am I right that a "Moved" account is actually a new account, with some data migrated and the old account redirecting to new?  @unattributed @Gargron @atomicpoet @mastodonusercount Yes, instance discovery needs to be through some form of spidering, perhaps users' following and follower lists.
 @Gargron @atomicpoet @mastodonusercount |
@Gargron @mastodonusercount Does your algorithm count instances that no longer exist?